By Dan Satherley
Imagine Parliament with the usual 120 MPs, except that 110 of them belong to Labour. Sounds ridiculous?
Well in 2002 that's what could have happened if New Zealand hadn't adopted MMP, according to a new election calculator developed by researchers at the University of Auckland.
In that year's election, Labour romped home with over 42 percent of the vote – but with 52 seats, still had to cut agreements with United Future and the Progressives to form a government.
National limped in with 27, but it could have been so much worse.
"Quite extreme results are possible under first-past-the-post," says Dr Geoff Pritchard, from the university's Department of Statistics, who developed the calculator with Dr Mark Wilson of the computer science department.
The 2002 general election provides an extreme example of what can happen when electoral systems don't allocate seats based on the proportion of votes.
It doesn't matter if the number of MPs is 120, as it is now, or 99, as it was under FPP - the point remains the same.
At this year's election – which some are predicting could be just as lopsided, but in favour of National – voters will be asked if they want to keep MMP, and if not, what alternative system they would prefer.
MMP, which New Zealand has used since the 1996 election, stands for mixed-member proportional. Quite simply, it allocates seats to parties based on the total percentage of the vote they receive – though there are quirks in it, such as the 5 percent threshold and the oft-maligned rule that allows parties with less than that to enter Parliament if they win a single electorate.
If we decide to keep MMP these idiosyncrasies will be up for review, but if we decide to throw the baby out with the bathwater, there are four other systems on offer – preferential voting, single-transferable vote, supplementary member and of course, good old first-past-the-post.
Dr Pritchard says there isn't enough awareness of how changing the electoral system can dramatically alter the balance of Parliament.
"A few people are perhaps aware of it in a general sort of way, but what we wanted to do was make it a bit more quantitative, and let people see some actual numbers and how things would work out under these different systems," says Dr Pritchard.
"There really are some quite wide variances possible, especially under some of these systems. Maybe people were more aware of that back then when it was more common to think of electorate boundaries being important and so on, and MMP has made some of those things a bit less important, so perhaps there's not as much awareness these days."
The variances typically happen when a party's core support is concentrated in a smaller number of electorates than that of their opponents.
"It used to matter quite a lot whether a particular town or even a particular street was in one electorate or another," says Dr Pritchard, "because that could tip the result one way or the other in a marginal electorate.
"Of course these days, that matters much less – most of the electorates we have now are just convenient administrative units, and only in a few cases does it actually matter who wins the local seat."
It also can occur when two parties appeal to the same constituency, splitting the vote, allowing a third party to triumph. This happened to Labour in 1978 and 1981 – for two consecutive elections it outpolled National, yet was unable to form a Government because Robert Muldoon's party was able to secure more electorates.
In those two elections, Social Credit obtained 16.1 and 20.65 percent of the vote, yet never managed to win more than two seats.
"People have their own ideas on how fair that sort of thing is," says Dr Pritchard.
"You can think maybe that's a good outcome or a bad one depending on how you view these things, but I suspect there probably isn't quite enough awareness with a moderately lopsided vote, quite extreme results are possible under first-past-the-post."
The calculator lets you punch in your own numbers – from an opinion poll, for example – or look at historic elections dating back to 1993.
The STV and preferential systems, which require information not recorded in past votes, are calculated using data from the 2008 New Zealand Election Study, in which participants rated political parties on a scale of zero to 10.
"That not only tells you which party they like the best, but also what their second and third and so on choices might be," says Dr Pritchard.
The other difficulty with reimagining the past is that not only have electoral boundaries shifted, but so has the demographic make-up of the country.
"It's not an exact rerun of those older elections in the 1990s, but more like if we had a pattern of first preference support like that now, what would happen," says Dr Pritchard.
"It's more like a simulation of a modern election in which the party support runs along 1990s lines, rather than an actual simulation of a 1990s election."
For politics and statistics junkies it's fascinating to look at what might have been – one example being the Parliament with 11 Labour MPs for every National MP.
"One thing I like to do is after you've put in a scenario, try exchanging the National and Labour support – make the two main parties change places – and run it again, and see what the result is then," says Dr Pritchard.
"That sort of gives you an idea whether there's a bias towards one party or the other. Do they in fact just exchange numbers of seats, which would be I suppose the perfectly fair way to do it, or do you get some different outcome?"
Dr Pritchard says the preferential systems tend to help the left.
"You can often see Labour doing a lot better because it can pick up second preferences from the Greens."
On the other hand, Dr Pritchard's colleague Dr Wilson notes supplementary member – the alternative system backed by the 'Put MMP to the Vote' campaign – delivers results very similar to FPP.
"With the simulator you can certainly see that it usually delivers a comparable number of seats in Parliament," says Dr Wilson.
In fact, the 2005 election would have been won by National under a supplementary member system.
After capturing the support of Winston Peters' NZ First by winning two more seats than the Don Brash-led National, Labour managed to scrape together a government. But under supplementary member, the calculator suggests National would have won two more seats than Labour.
Kiwiblog's David Farrar last year showed how Dr Brash could have formed the government with the support of the Maori Party and United Future – without the need for party he now leads, ACT.
No wonder Labour leader Phil Goff has accused the anti-MMP campaign of being a front for right-wing business interests.
"Perhaps they feel that the political system is easier for them to control if they've got a first past the post system rather than a multi-party system," he told TVNZ in June.
Still, the hypothetical 2002 election results show that when a party is really, really unpopular, it will usually lose – no matter the electoral system.
But as Dr Pritchard notes, with non-proportional systems like FPP, all it takes is a little imagination, and anything can happen.
"Had we drawn the electorate boundaries in a different way, we might have got a different result."
3 News